Two-stage optimal scheduling of an islanded microgrid considering uncertainties of renewable energy

Scheduling islanded microgrids in a reliable, economical, and efficient manner is challenging due to the strong uncertainty and randomness of renewable energy generations, like photovoltaic and wind sources. Chance-constrained programming methods have been proposed to balance power supply and demand, but these often consider known probability distributions and chance constraints separately, leading to suboptimal solutions. To address this limitation, this study proposes a novel distributionally robust joint chance-constrained program for modeling the two-stage energy and reserve economic scheduling problem of an islanded microgrid. The Wasserstein distance is used to capture the random characteristics of photovoltaic sources. An optimized Conditional Value-at-Risk (CVaR) approximation method is applied to simplify the conic program of the model into a computationally tractable linear program. Finally, the case study validates that the proposed method reduces solution conservativeness compared to the combined Bonferroni and CVaR approximation method, which considers chance constraints individually. The proposed method enables efficient and practical scheduling decisions for islanded microgrids by considering the joint chance constraints and the uncertain nature of renewable energy.